We're at Fintech Meetup Europe in Lisbon from 6th to 8th October. 🇵🇹 Our CEO @dimitrimasin joins the day 1 panel "Replace or Reshape? Designing the Workforce of the Future” to talk about what actually changes for customer operations teams when AI agents take on the work. We've also got meeting slots across the three days for banks, insurers, and fintechs looking at automating frontline and back-office work. Drop a comment or DM us to book one. 🤝

We spent last week in Lisbon for our first team offsite. 🙌 This was a great opportunity for our new joiners to meet the old crew, connect in person with the teams spread across time zones, and align on the exciting goals that will shape our strategy in 2027. Highlights from the week included: 🦾 A hackathon, where small groups brainstormed innovative new capabilities for our agents or automated a process dragging down the day to day. 🏄 Surf class, tile painting, and a morning workout led by our very own Ashley Rosenthal. 📈 Long-term vision alignment sessions from leadership, paired with practical workshops to help teams hone in on their goals and priority projects. 🌊 Seaside walks, bonfires, and yoga. We’re back in our regular offices feeling more connected as a team and fired up with new ideas to take us to 2027 and beyond. 🚀

Last week we hosted our first Gradient Collabs session in London with @yondercard! Antony Atkins, Senior Escalations Manager at Yonder, sat down with Yulia Yan from our AI Delivery team to talk about Yonder’s two year journey running our AI agents in production, most recently to automate disputes. A few things that stayed with us: 🧱 A safe rollout is a team sport. Ops, risk, and engineering had to agree what "good" looked like before anything went near a customer. 🧠 The mindset change is harder than the technology. Teams move from clearing a queue to supervising the work, and that takes time. 🤝 Start with the process you already know is broken. Not the one that's easiest to automate. Then everyone sat down to eat, and the conversation carried on for another two hours. Thank you Antony and the Yonder team, and thank you to everyone who joined us around the table. 🧡

You can now find Gradient Labs listed as a supplier member of the Credit Union Executives Society (CUES). 🎉 Our team came together through relationships built over a decade in UK fintech, so we understand the importance of community and building something for people who trust you. 🤝 We also know our AI agents make the biggest difference to lean teams carrying more volume than headcount. At Zego, our AI agent scores a 77% customer satisfaction rating against 61% for their human agents, and we hold 90% CSAT on disputes at Yonder. That's why we're confident we can help credit union member experience teams move faster, without losing the human touch that members actually notice. If you're in the community, come say hi. 👋

Last week we had a full table at @DanielBoulud's Le Pavillon in New York. 🗽 We brought together operations, product, and technology leaders from banks, credit unions, lenders, and fintechs on the evening of @dimitrimasin's session at @finovate Fall. The conversation: what does it actually take to run AI inside a regulated support team? More of these coming. If you run customer operations at a bank, lender, or fintech, drop a comment to join us at our next one. 🍽️

Our CTO @neal_lathia took the 'AI in Action' stage at @FintechWeekLDN last week, alongside Sean Hederman, CTO at Zilch, for a panel on moving from copilots to agents. Three things that came up in discussion: 🎯 Scope beats ambition. Agents earn trust on narrow, well-defined work long before they get near end-to-end ownership. 🛡️ Guardrails have to be deterministic. You can't rely on the model to police itself. 👥 Humans must stay meaningfully involved, doing real judgment work rather than approving whatever the agent suggests. That last point: this is the philosophy that’s allowed our disputes automation with Yonder to thrive. The agent runs the full case, 80% arrive fully evidenced, and decisions that used to take a week+ now take a day on average. Thanks to @fintechfringe for having us.

We’ve been named the #1 most talent-dense scale up in the UK by Nova. 🚀🇬🇧 Their Talent Density Score ranks every person currently working at a company on experience, education, international background, and languages, then averages that across the whole team. What a fitting recognition as we head into our first global offsite this week. 🤗 Also, we're hiring! ✨ Check the link in the thread. ⬇️

The team’s all here at Finovate Fall, ready to hear CEO @dimitrimasin speak on the main stage later this afternoon. 🙌 We’ll be in the city through the end of the week, so if you can’t make it to the event, feel free to reach out for a coffee or chat outside of the conference. We’ll take all excuses to walk around Manhattan and see more of our ads on the street. 🤭 Have you spotted one yet?

Manhattan, have you noticed us on your commute? ✨🚀

📣 Gradient Labs is launching its first-ever Gradient Collabs series, where we sit down with one of our customers to hear their real-world insights from working with AI agents. 🤝 Our guest coming up next week is Antony Atkins at @yondercard, in conversation with our very own Yulia Yan. We’ll be hosting Antony and senior leaders across banking and fintech at an intimate, invite-only dinner in London, where we’ll dive into Yonder’s impressive work on disputes automation. For a sneak peek on what they’ll talk about, check out Yonder’s case study: eu1.hubs.ly/H0xT0y90 Interested in joining us for this event? Leave us a comment ⬇️

When you dispute a charge on your card, the worst part is the wait. In most cases, it takes at least a week to find out if you’ll get your money back. @yondercard, the UK fintech behind lifestyle reward cards, has ended that wait with Gradient Labs. Dispute cycles that used to take a week now close in 1 to 2 days, even during volume spikes. Because our frontline and back-office agents share the full context of a case, 80% of disputes arrive for human review fully evidenced the first time. This empowers faster decisions for human teams who start with all the information up front, as opposed to having to chase context with a customer days later. What do Yonder’s members think about the new system? Customer satisfaction ratings have jumped to 90%. 🧡 Read how Yonder did it: eu1.hubs.ly/H0xJxbd0

Our team spent a few days in Atlanta last week at Fintech South, where we also teed up for some Topgolf in between conversations with senior leaders in banking, payments, and fintech. ⛳ Our biggest swing of the conference? Explaining that our AI agents have already sailed far past the pilot stage and into deployment at scale for banks and fintechs worldwide. Disputes? Automating 150% faster. CSAT? Outpacing human teams. 🦾 All par for the course over here.

See you real soon, NYC. 🗽 We’ll be at Finovate Fall, and our CEO @dimitrimasin will be on the mainstage for a panel with Royal Bank of Canada, Prudential, and the University of Michigan Credit Union. The question on the table: as agentic AI moves past live pilots, where are financial institutions actually making or saving money? 💰 If you’re in NYC the week of September 9th, drop us a line! Our team on the ground would love to connect.

So the can agent improve itself, with your supervision. It proposes fixes from what it sees in production, and makes every failure a permanent test case to prevent future regressions.

Today we're launching Collaborate: edit your AI agent like code, in the tools your teams already know and love. Learn more below or at gradient-labs.ai/product/col…

@yondercard used to need 5-7 days to close each disputes case. Now these cases are resolved in 1-2 days. The major UK fintech automated disputes by connecting our frontline and disputes AI agents so a single case keeps its full context, from first message to final decision. Senior Escalations Manager Antony Atkins put it this way: “We've used plenty of AI tools, and nothing else has come close to this, especially for disputes. It helps that Gradient Labs genuinely knows financial services; we're working with people who understand disputes as well as we do.” Disputes now close 150% faster, 80% of cases arrive fully evidenced, and CSAT holds above 90% through the busiest weeks. Read how they did it in the thread. ⬇️

🎙️ Join us next Thursday for a free webinar on how agentic AI can make customer operations safer, more efficient, and more productive… even in a highly regulated financial institution. We’ll be hosted by @khaslett, and our very own @neal_lathia will talk about: 🔒 How to bring compliance and security stakeholders in early. 🎯 Why starting narrow beats going broad when it comes to agentic AI. 🧪 How offline testing builds the confidence needed to move from pilot to production. 📊 How to measure an AI agent’s performance and impact. This webinar is a working session for executives to think about where in their institution AI agents can reduce cost and resolution time without adding operational risk. Register to join us: eu1.hubs.ly/H0xvPK50

We had a full house in our office last week of founders and engineers across AI, backend, and infra. The discussions and networking were led by technical leaders Hayyaan Ahmad of Round Treasury, @nfcampos of Witan Labs, @emilsjolander of Legora, and our very own CTO @neal_lathia, who each presented their learnings on what it’s really like to build AI inside an AI-native company. Thanks to all who stopped by!

One of our favourite things to do is get senior leaders together to have meaningful conversations about the future of AI in financial services. Last week in NYC, we took to Soho to do just that. While rain kept us off the rooftop (and away from the Manhattan views), it didn’t stop the lively conversation about AI and what it really takes to see ROI from agent deployments in a regulated industry. Thanks to all who joined. 🎉

Our team members Mark Kutz and Ashley Rosenthal will bring Gradient Labs to Atlanta, Georgia in two weeks. 🍑 They’ll be at @FintechSouth, ready to show what our agents resolve in production: support, disputes, collections. It's our first event in Transaction Alley, the city 70% of US card payments run through. If you’re going and want to grab 20 minutes with us, drop us a comment. ⬇️

KYB is one of the most document-heavy steps for a bank trying to onboard a new business. Certificates of incorporation, proof of address, ownership structures, and director details all arrive in different formats, and someone has to read each one, check it against policy, and chase whatever is missing. Our AI agent runs the whole KYB check. It reads every document a business submits, verifies registration, directors, address, and ownership against your internal policy, screens against sanctions lists, and emails the customer for anything missing or out of date. Verified businesses route straight to onboarding, and only the genuine edge cases reach a human. Every check and decision is logged for audit. 📋 Find out more about how this works at the link in the thread.

Our team members Ashley Rosenthal and Zan Faruqui are back on the east coast after an exciting few days in Chicago at the Midwest Acquirers Conference, where they spent time with payments and banking leaders talking about where AI actually earns its place in customer operations, and where it doesn't. They also got the chance to experience Alinea, a definite 2-star Michelin recommendation for anyone in the Chicago area. ⭐⭐

It happens frequently: new customers sign up for an account or credit card, get partway through onboarding, then disappear. Maybe they're confused on a verification step, hesitating over a first deposit, or have questions about terms and perks. Our AI agent reaches out the moment a customer misses a milestone you define, like a first deposit within 7 days or verification within 3, on the channel they prefer. It asks what's holding them back, walks them through the specific step they're stuck on, and either resolves it in the conversation or routes the case to your team if it needs human judgement. Your customers get proactive help, you get an immediate uplift in early month on book activation, and your team only sees the cases that genuinely need them. 💪  Read more at the link in the thread.

Next week, four engineering leaders will share lightning talks on what they’ve built, diving into the infrastructure behind the AI agents that actually do the work. Join us on 30th July in London to hear from: ⚡ @nfcampos (Co-founder & CTO, Witan Labs, creator of LangGraph) ⚡ @neal_lathia (Co-founder & CTO, Gradient Labs) ⚡ @emilsjolander (Member of Technical Staff, Legora - previous Director of Eng at Figma) ⚡ Hayyaan Ahmad (Co-founder, Round Treasury) There’ll be drinks, networking, and a lot of engineering conversations. RSVP at the link in the thread.

The customers we work with test our AI agents rigorously before anything goes live to their end users. One team runs 60+ manual tests before launching any new procedure. We listened to customer feedback to make this internal testing process even faster, with two key updates to our testing environment that we built and deployed in just one week. 🦾 Now: almost no setup time to test a full use case end to end in our platform, which gives teams the speed and ease they need to roll out the AI agent across more complex customer operations work, with full confidence.

We interrupt our regularly-scheduled content to bring you… the newest puppy in the office. 🤩 While Loaf has spent decent time on brand duty, Reece hasn’t figured out how to wear the “good bot” hat yet. 🤭 Want to join our team in London and meet these four-pawed friends IRL? We still have open roles. 🎉  (Link in the thread.)

Did you catch us on @AnthropicAI's website? 🎙️ We’re featured in a customer story about how our team uses Claude to automate customer operations for banks, fintechs, lenders, and more. Most AI support tools stop at the simple queries. This piece covers why we went after the complex, regulated work financial institutions actually escalate across frontline and back office systems, and why Claude's natural, accurate responses made it the right model to build on. When the bar is 80-90% resolution rate, the right technology matters. Link in the thread. ⬇️

🗽 New York, want to swap notes on how AI agents are actually working for your business? We’re gathering a group of leaders in banking, fintech, and insurance for private rooftop drinks in Soho and casual networking and conversation. Topics that will be on our mind? AI implementation, what’s working for real teams out there (or not), and how your program can actually move the needle. If you’d like to join us, check the link in the thread for more information. ⬇️

🚨 Four engineering leaders, 12 minutes each. 🚨 If you’re an AI engineer, backend, infra, or founder building with agents, join us on 30th July in London to hear lightning talks from: ⚡ @nfcampos (Co-founder & CTO, Witan Labs, creator of LangGraph) ⚡ @neal_lathia (Co-founder & CTO, Gradient Labs) ⚡ @emilsjolander (Member of Technical Staff, Legora - previous Director of Eng at Figma) ⚡ Hayyaan Ahmad (Co-founder, Round Treasury) There’ll be drinks, networking, and a lot of conversations about what these teams have built that they didn’t think was possible before. RSVP at the link in the thread.

Watch the full video here: piped.video/ICZLiV96giI

Few companies are as focused on customer experience as @Wise. We sat down with Rohan Basu, Global Head of Operations, and @jerganc, Product Lead for Customer Support, to discuss how AI - and Wise’s partnership with Gradient Labs - is helping customers get faster, clearer support at scale. The conversation covers resolution over deflection, quality at scale, and the role of human teams in the moments where trust matters most. Watch the full video with @dimitrimasin in the thread. 👇

Always-on, instant support is the baseline customers expect from their financial institutions. Banks are still catching up to expectation, and a foreign time zone is where the gap shows most. We surveyed two thousand travellers about banking problems abroad, and we found: 🌙 7% of incidents happened overnight, with no one reachable to fix them ⏳ Nearly 1 in 4 travellers waited 30 minutes or more to reach their bank 📵 6% never got through at all Every minute on hold in a foreign country is a minute the customer spends reconsidering their bank, often with unexpected call costs and sightseeing plans derailed. A capable AI agent answers the moment a customer is stranded, in any language, at any hour. It gathers the details, reassures the customer, and resolves the common cases outright, leaving human agents free for the truly tricky ones with full context already attached. Even when a human is needed, the AI agent has already gathered every detail and reassured the customer. The case is ready to resolve the moment someone's shift begins, instead of waiting for business hours to come round. The banks that make fast resolution the rule are the ones these travellers keep. Read more in the thread. 👇

Our CEO @dimitrimasin sat down with @peterrenton to talk about what it really looks like to have an AI agent live in your day-to-day banking operation. The highlights: 🔒 Autonomous agents beat co-pilots when it comes to quality and compliance over time. Handing work over to an autonomous agent is actually less risky (tune in to hear Dimitri explain why). 💡 AI agent performance should be benchmarked against human standards, not theoretical perfection. 💪 None of Gradient Labs’ customers have laid off human staff because of our AI agents. When AI is doing its job well, it’s automating work that actually saves teams time and meets a real business gap. This empowers human teams; it doesn’t make them irrelevant. Check the thread to learn more. ⬇️

Picture a customer at a foreign checkout. Their card won't work, it's 11pm, and their bank's support line is closed back home. What happens then? We surveyed two thousand travellers about exactly this. Half of them had hit a card, payment, or banking issue abroad in the past two years. This is what it costs travellers: 💳 40% were left more than $200 out of pocket, and 13% more than $500 🏨 22% slept somewhere unplanned; 18% described themselves as "stranded" 📉 24% had to borrow money or go into debt because of a banking issue The cost to the bank is even steeper. 39% of these travellers switched or seriously considered switching after a bad support experience abroad, and 35% warned friends and family away from a specific bank. These failures are avoidable, some easily so, with AI agents purpose-built for finance. Full survey in the thread. 👇

Only two years ago, most banks and fintechs we spoke to were standing up their first AI support agent. Every question was about that one agent: can it answer well, make tool calls safely, and follow processes faithfully? Along the way, AI became a genuine lever to improve the customer experience, and many of our customers hit CSAT numbers they didn’t think were possible. 📈 💡 But now, our most forward-thinking customers are planning something bigger: their multi-agent future. Customer-facing support isn't where most of the handling time goes in finance. Back-office work dominates an operations team's day, from disputes and KYC to collections and fraud investigations. The teams furthest ahead have stopped asking how to run one agent well. They're architecting many that work together to automate cross-team processes end to end. ⚙️ Our new design lets financial services build multiple specialist agents that run as one system, sharing context and passing work between each other where a customer's journey requires it. There are huge time and cost-saving implications. For example: chaining the frontline and disputes agent together cuts investigation times from days to minutes. ⏱️ The reason our customers can build this today comes down to the platform underneath: multi-agent orchestration, finance-native guardrails running on every turn, an audit trail that holds up to regulators, and context that carries across channels and days. 🦾

Check out the @OpenAI homepage today. You'll find Gradient Labs there. We've deployed AI agents for some of the largest, most regulated financial institutions in the world. There is no room for error. Our agents must follow strict procedures, maintain state across interruptions, and respond quickly enough for voice while maintaining compliance. We use OpenAI models because they meet those standards: high instruction accuracy, low hallucination rates, and ~500ms latency in production. This results in safe, compliant conversations that sound as natural as a real human agent, even on voice. Read the case study here: openai.com/index/gradient-la…

Last week, we turned three! 🎉 We celebrated in London with the customers and community who got us here. Three years ago Gradient Labs was an idea about what AI agents could do for financial services. Today it's used by banks, lenders, and fintechs around the world to automate critical customer operations and improve the customer experience at the same time. Thanks to everyone who turned out, and to everyone who's backed us since year one. Here's to many more. 🥂

Last week we shared that everyone at Gradient Labs, from Engineering to Sales, uses Claude Code to build their own skills and automations. Some of you asked for specific examples. So this is the Marketing team, stepping in to share one of our hacks. 🦾 As marketers, we get a constant stream of event sponsorship invites. We built a Claude automation that combs our inboxes, adds each new event to a database, and writes a pros and cons list for every one, along with a product-market fit analysis and a cost estimate. We review the database regularly, do our own research, and make participation decisions from there. It saves us significant time each week and keeps all our information in one place. Essential for a lean team running a global operation. 📈

When it comes to guardrails, we prefer to show you the numbers. Ours run on every turn and protect both sides of the conversation. Customer guardrails detect complaints, vulnerability, and financial difficulty, then hand off or reroute. Agent guardrails catch tipping-off, disallowed terminology, and advice the agent can't legally give, editing the draft before it ever reaches the customer. And now you can see how well each one is working. Click on any guardrail in the Gradient Labs platform, and you'll see 90 days of performance history: how often it ran, how often it handed off, and its Precision, Catch rate, and Accuracy. Your compliance team gets evidence instead of assurances. So do you. 🦾

From 23-25 June we’ll be at Celtic Manor for Credit Week, conversing with the UK’s credit and lending community. On 24 June, our Chief Scientist @danaintoniou will host 30 senior lending leaders in a Solution Room session to work through the operational problems slowing their teams down: customers dropping off at sign-up, manual headaches in underwriting, securing promises to pay from customers in early arrears, and more. If you’ll be there, come say hi at booth 35. 👋

Last week in Amsterdam, we hosted a dinner for banking and finance leaders attending Money20/20 Europe. It was a great conversation with a few recurring themes: 🎯 C-suite caution can be overcome by meeting a clear bar. Fast ROI and measurable time and cost savings are critical proof points, so selecting the right problem to solve is arguably more important than fast implementation. 🔧 Teams aren’t looking to overhaul their existing processes. They want AI that works with current tech stacks, that can slot alongside how teams currently operate, and that non-specialists can manage day to day. ⚡ Every bank in the room could name something they'd struggled to scale for years. That's where AI delivers the clearest, fastest return. Thanks to everyone who came, and to the team at Sinne for a fantastic meal.

We’re thrilled to announce that @current, a U.S.-based consumer fintech banking platform with over six million members, is partnering with Gradient Labs on their AI agent for customer operations. Current’s mission is to improve financial outcomes. They chose us as their partner because we've proven AI can improve those outcomes and lead to meaningfully stronger customer experiences. As the only AI agent purpose-built for financial services, Gradient Labs is helping Current deliver the kind of 24/7 support that is essential for building and maintaining customer trust.

POV: a week at @money2020 Europe in Amsterdam 🇳🇱 If you missed the key insights shared by our CTO @neal_lathia and Head of Marketing @emmalynefurlong this week, scroll back through our posts to read their takeaways. 📈 #Money2020EU

It's @emmalynefurlong here on the ground at @money2020 Europe, sharing a few reflections after three days of keynotes and conversations on the floor. Leaders across financial services are (rightly) adamant about maintaining their customers’ trust as they evaluate AI. I thought Amit Thawani, CIO of Insurance at @LBGplc, articulated this particularly well on the Engineering an AI native Bank Panel, where he said: When you’re working in high-stakes, emotional life moments like bereavements, you cannot afford to get the experience wrong. AI can reduce the operational burden, but the line between success and failure is in the customer experience. Empathy and CSAT matter just as much as efficiency. But the need for efficiency is not to be discounted, and was clearly top of mind at the show. Nearly everyone we spoke to at our booth had a specific, manual back-office process they’re struggling to automate. It seems the discourse is shifting from “How can I use AI?” to “I’ve outlined the biggest challenges in my business that are ripe for automation, now how do I use AI to solve those challenges safely, and without compromising customer experience?" It’s been validating to see the similarity in approach between how we run Gradient Labs, and the companies we serve: that any AI project must start with a clear business challenge, eliminate painful manual processes, and make the customer experience fundamentally better, all at the same time. #Money2020EU

The real MVP of the conference. ☕️ #Money2020EU

@neal_lathia here, live from @money2020 Europe. Earlier this week we shared our increased $26M Series A, and then I landed straight on the conference floor in Amsterdam. Two days in, this is what I keep hearing: The build vs. buy question dominates every AI conversation in finance right now. My honest take is that it’s the wrong question to answer first, and possibly the wrong frame altogether. Every team I’ve spoken to that has gone down the internal build route has ended up with something narrower than planned, shipped later than expected, and still needing to wrap around things they had not anticipated. Whatever you build, you are building on top of something. Whatever you buy, you are building on top of that too. The conversations that go somewhere start differently: what are you actually solving for, and what does “good” look like for your business? That sounds obvious, but it is genuinely rare, and the teams making real progress are the ones who can answer that question precisely. Once you know what you’re looking for, you can test internal models against agents out there and see which performs best for your needs. Of course, I also stopped by the LHV Bank booth to say hi. Always good to see customers here in the wild. #Money2020EU

Who remembers these robot dogs from 2000? We got this one off eBay for our booth at #Money2020EU and had a laugh at the description in the manual:

Six months ago, Pockit came to us with two challenges: 1️⃣ Their legacy chatbot was deflecting tickets, but CSAT kept slipping. Customers didn't feel their problems were actually solved. 2️⃣ As Pockit scaled, support volume spiked in ways that were hard to plan for. Today, more than 7 in 10 inbound queries are resolved end-to-end by Gradient Labs’ AI agent, and customer satisfaction is 80%. The team did it without adding headcount. Michiel Smet, Head of Operations at Pockit, shared: "We truly think that if people have a problem and you solve it, that builds brand loyalty. That's why customer resolution is so important. With Gradient Labs, we have an AI agent that's actually resolving problems, boosting our CSAT rating, and absorbing growth.” Read Pockit’s story in the thread.

Our “good bot” team is live on the ground at @money2020 Europe. 🇳🇱 If you’re not on the conference floor with us, don’t worry: we’re about to bring it to you. For the next two days, our CTO and co-founder @neal_lathia and our Head of Marketing Emma Martin will take over the Gradient Labs account to bring you real insights and observations gained on the go at Money20/20. You can follow along here and on LinkedIn. And if you’re with us in Amsterdam, stop by our booth (5D58) to say hi and play with our retro robot dog. 🐕🤖 #Money2020EU

ICYMI: we’re one of eight startups selected for @NatWestGroup's 2026 Fintech programme. 🏦 This year's theme is "How is AI shaping the future of customer experience?" Candidates needed to prove product-market fit, an active user base, and real revenue. They also needed to prove their product meaningfully changes how financial services companies handle the customer experience. Over the next 12 weeks, we’ll be partnering closely with NatWest’s Open Innovation team, exploring collaboration opportunities to advance customer experience with specialised AI agents across customer operations. Thrilled to be part of this programme. Stay tuned for what’s next.